WANG Yuan-zheng ,Olusegun lDOWU ,WANG Yun ,HOMMA Koki ,NAKAZAKl Tetsuya ,ZHENG Wen-jingXU Zheng-jin#,SHlRAlWA Tatsuhiko

1 Rice Research Institute,Liaoning Academy of Agricultural Sciences,Shenyang 110101,P.R.China

2 Division of Agronomy and Horticulture Science,Graduate School of Agriculture,Kyoto University,Kyoto 606-8501,Japan

3 Rice Research Institute,Shenyang Agricultural University/Key Laboratory of Northeast Rice Biology and Breeding,Ministry of Agriculture and Rural Affairs/Key Laboratory of Northern Japonica Super Rice Breeding,Ministry of Education,Shenyang 110866,P.R.China

4 Division of Biological Resource Sciences,Graduate School of Agricultural Science,Tohoku University,Sendai 980-8576,Japan

Abstract The dense and erect panicle (EP) genotype conferred by DEP1 has been widely used in the breeding of high-yield Chinese japonica rice varieties. However,the breeding value of the EP genotype has rarely been determined at the plant population level. Therefore,the effects of the interaction of EP genotype and the environment at different locations and times on rice yield and its various components were investigated in this study. Two sets of near-isogenic lines (NILs)of EP and non-EP (NEP) genotypes with Liaojing 5 (LG5) and Akitakomachi (AKI) backgrounds were grown in the field in 2016 and 2017 in Shenyang,China,and Kyoto,Japan. In 2018,these sets were grown only in Kyoto,Japan. The average yields of the EP and NEP genotypes were 6.67 and 6.13 t ha-1 for the AKI background,and 6.66 and 6.58 t ha-1 for the LG5 background,respectively. The EP genotype positively affected panicle number (PN) and grain number per square meter (GNPM),mostly resulting in a positive effect on harvest index (HI). In contrast,the EP genotype exerted a negative effect on thousand-grain weight (KGW). The ratio of the performance of the EP genotype relative to the NEP genotype in terms of yield and total biomass correlated positively with mean daily solar radiation during a 40-day period around heading. These results indicate that the effectiveness of the EP genotype depends on the availability of solar radiation,and the effect of this genotype is consistently positive for sink formation,conditional in terms of source capacity,and positive in a high-radiation environment.

Keywords: erect panicle,genotype by environment interaction,rice (Oryza sativa L.),solar radiation,yield

1.lntroduction

The firstjaponicaerect panicle (EP) variety,Liaojing 5(LG5),was developed in China in 1976 using the Italianjaponicavariety Balilla as the genetic source (Xuet al.1995;Sunet al.2012). Since then,LG5 and its derivatives have been used as donors of the EP gene in the breeding of high-yield rice varieties in China. The EP gene was subsequently mapped and cloned into the EP-type rice,LG5,and Qianchongliang 2 (Huanget al.2009;Wanget al.2009). The EP genotype originates from a gain-of-function mutation in theDEP1(EP/qPE9-1) gene,resulting in the truncation of the phosphatidylethanolamine-binding protein-like domain protein,a significant enhancement in meristematic activity,reduced inflorescence internode length,and increased grain number per panicle (GNPP),which ultimately result in an increase in yield (Zhouet al.2009;Sunet al.2014;Xuet al.2015).DEP1markedly enhances grain yield primarily by increasing the number of secondary branches and the number of grains on each secondary branch(Xuet al.2010). The EP gene remarkably increases the grain density and the number of filled grains per panicle(Ashikariet al.2005;Xuet al.2014). Moreover,EP rice cultivars tend to have high lodging and fertilizer resistance because of their characteristic plant canopy structure(Xuet al.1990,2004;Qiaoet al.2011). These cultivars also effectively use solar energy,show accelerated CO2diffusion,and exhibit improved ecological growth conditions in the middle and lower parts of the rice canopy(Xuet al.1995;Tanet al.2001).

The effects of plant type vary among the four regions in China (Jinet al.2013),although no direct evidence has confirmed the interaction between genotype and environment. Ecological conditions may influence the effects of plant type on yield and yield-related traits. A comparison of EP rice and other high-and moderateyielding cultivars (Hirookaet al.2018) suggested that growth duration may affect the growth performance of EP rice,although this hypothesis has not been fully tested.

Understanding the effects of the ecological environment on the effectiveness of the EP genotype in terms of yield and yield-related traits will provide useful information for breeding rice ideotypes and optimizing management practices. However,the interaction between EP genotype and the environment has rarely been investigated at the plant population level (Wanget al.2009). Moreover,several of the relevant studies have used varieties other than NILs (Jinet al.2013;Lianget al.2015;Tanget al.2017;Hirookaet al.2018) or employed limited plot sizes rather than plant populations(Jinet al.2013;Sunet al.2014). Determining whether the effect of the EP genotype is consistent across different conditions remains challenging because of the effect of its genetic background on grain yield and the lack of population-based studies that have assessed the effect on plant stature. Moreover,the factors affecting the EP genotype should also be determined. Therefore,the effects of interactions between the EP genotype and the environment on rice yield and factors influencing yield were investigated in this study.

2.Materials and methods

2.1.Plant materials

In this study,four NILs were developed using markerassisted selection with 550 pairs of single-sequence repeat (SSR) molecular markers distributed across 12 chromosomes of the rice (OryzasativaL.) genome(Konget al.2007). Two hundred recombinant inbred lines (RILs) (F7) were developed from a cross between the first EP variety to be widely used in northern China,LG5,which carries the EP alleledep1,and the curved panicle-type Japanesejaponicacultivar Akitakomachi(AKI),which carries the curved panicle alleleDEP1.From these 200 RILs,a RIL named R87 that contains the homozygousdep1allele from LG5 and 82.6% of the genetic background of AKI was selected and backcrossed with AKI three times. Two successive self-pollinations of BC3F5plants heterozygous for theDEP1fragment resulted in homozygous NILs named AKI-EP withdep1and AKI-non-EP (NEP) withDEP1,which contained most of the genetic background of AKI except for thedep1orDEP1introgressed segment. The method described above was then used to construct LG5-EP withdep1and LG5-NEP withDEP1,which carry thedep1andDEP1regions from LG5 and AKI,respectively,and have most of the genetic background of LG5. The AKI-NEP and LG5-EP lines were not the original parents,but they were selected from the backcrossed population.

2.2.Plant culture

The four NILs were grown using conventional crop management practices in Shenyang (China) in 2016 and 2017 and in Kyoto (Japan) in 2016,2017 and 2018. The experiment was conducted at the Experimental Farm of Shenyang Agricultural University (41.8°N,123.4°E) during the rice-growing season. The germinated seeds were grown,and the seedlings were transplanted into the field.The sowing dates were April 21,2016 and April 14,2017,and the transplantation dates were May 26,2016 and May 26,2017 in the first and second years,respectively. For the study at the Experimental Farm of Kyoto University(34.0°N,135.8°E),the sowing dates were April 22,2016,April 20,2017,and April 20,2018,during the first,second,and third years of the experiment,respectively,and the seedlings were transplanted on May 19,2016,May 16,2017,and May 18,2018,respectively.

The four NILs were grown with 15 cm×30 cm spacing(22.2 plant m-2),and nitrogen fertilizer was applied at 6 g m-2as a coated urea,along with P2O5(6 g m-2) and K2O (10 g m-2). A randomized complete block design was used for each NIL. The agronomic practices used conventional methods,such as irrigation and control of weeds,insects,and diseases,which were maintained as required to prevent pest infestation and avoid yield loss.

2.3.Measurements

The experiments included three replicates,each with 90 plants,all of which were observed,and the heading and maturity dates were recorded. The heading date is the date when more than 80% of the plants have safely completed heading,and the maturity date is when the rice husk turns yellow,the rice grain moisture decreases,the dry matter weight reaches a fixed value,and the grain becomes hard and difficult to break. The samples were collected at maturity. Fifteen representative plants were sampled from the middle of each plot when 80% of the plants reached physiological maturity to measure yield and yield components: biomass of aboveground plant materials,harvest index (HI),panicle number (PN),grain number per panicle (GNPP),grain number per square meter (GNPM),seed-set rate (SSR),and 1 000-grain weight (KGW). Plant height (PH) was measured during the full heading stage using five selected plants in the middle of the field from the sampled plant population.Whole plants were dried naturally post-harvest and these plants were stored in the greenhouse for at least one month before the determination of the yield and yield components. Data on climatic factors (daily air temperature and solar radiation) were collected from the meteorological weather stations adjacent to the research fields in 2016 and 2017 for Shenyang and in 2016,2017,and 2018 for Kyoto.

2.4.Definition of the effectiveness of the EP genotype

The effectiveness of the EP genotype was estimated as the ratio of EP genotype performance to the NEP genotype for each background,as follows:

Effectiveness=EP genotype/NEP genotype

when effectiveness is 1,＞1,or ＜1,it is considered to be neutral,more,or less effective,respectively.

2.5.Statistical analysis

All statistical analyses were performed using Microsoft Excel and R Software (R Core Team 2015). Analysis of variance (ANOVA) was performed using R Software,and the graphs were created using Excel 2017.

3.Results

3.1.Temperature and solar radiation

The daily mean temperatures (TMEANs) measured from the day of heading (for early AKI) to the day of maturity (for the late LG5) for both the Shenyang and Kyoto regions in 2016 and 2017 are shown in Fig.1. For the AKI background,the TMEANs in Shenyang were 21.6 and 20.9°C in 2016 and 2017,respectively,which are lower than the respective values of 28.8 and 27.9°C in Kyoto.There were no significant differences in TMEANs between the two years at either location. For the LG5 background,the TMEANs were lower in Shenyang (23.7°C) than in Kyoto (25.1°C) in 2016 and 2017. This difference resulted in the maturity stage of the LG5 background lines occurring 25 to 28 d later in Shenyang than in Kyoto.However,at both locations,the TMEANs of the grainfilling period were lower in the LG5 background than in the AKI background. In 2018 in Kyoto,the cumulative TMEANs from the day of heading for early AKI to the day of maturity for late LG5 was 28.7°C (Appendix A).

Fig.1 Daily temperatures in 2016 and 2017 at Shenyang,China (A) and Kyoto,Japan (B),for the periods from heading of the early var.Akitakomachi to maturity of the late var.Liaojing 5.

The daily solar radiation levels from the heading of early AKI to the maturity of late LG5 in Shenyang and Kyoto in 2016 and 2017 are shown in Fig.2. For the AKI background,the daily mean solar radiation in Shenyang in 2016 was 17.3 MJ m-2d-1,which was significantly lower than the 19.7 MJ m-2d-1in Kyoto. For the LG5 background,the average solar radiation in Shenyang in 2016 was 14.2 MJ m-2d-1,which was lower than the 18.5 MJ m-2d-1in Kyoto. In 2017,the daily average cumulative solar radiation in Shenyang was approximately 16 MJ m-2d-1for the AKI and LG5 backgrounds,which was slightly higher than that in 2016 for the LG5 background. The daily average cumulative solar radiation in Kyoto in 2017 was approximately 17 MJ m-2d-1for both backgrounds,which was slightly lower than in 2016.In 2018,the average cumulative solar radiation in Kyoto was approximately 20.5 MJ m-2d-1for both backgrounds,which was higher than in 2016 (Appendix A).

Fig.2 Daily solar radiation in 2016 and 2017 at Shenyang,China (A) and Kyoto,Japan (B),for the periods from heading of the early var.Akitakomachi to maturity of the late var.Liaojing 5.

3.2.Growth duration and PH

The growth periods for the EP and NEP genotypes with the AKI and LG5 backgrounds in 2016 and 2017 for Shenyang and Kyoto are shown in Table 1. The growth periods differed among genotypes,backgrounds,and locations. Heading was reached 2-5 days earlier in the EP genotypes with AKI and LG5 backgrounds in Shenyang than in the NEP genotypes in 2016 and 2017.In Kyoto,the NEP genotype with an AKI background reached heading earlier than EP,andvice versafor the LG5 background.

Table 1 Growth duration for the erect pancle (EP) and non-EP genotypes with Akitakomachi (AKI) and Liaojing 5 (LG5)backgrounds in 2016 and 2017 at Shenyang,China and Kyoto,Japan

The transplanting to maturity period in Shenyang was 10-28 days longer than in Kyoto for the EP genotypes.The period from heading to maturity in Shenyang in 2016 was 7-11 days longer than in Kyoto,whereas there was no significant difference in 2017. The period from heading to maturity was 5 days longer in 2017 than in 2016 in both NILs of both backgrounds in Kyoto,presumably due to the low temperatures in 2017.

The mean PH values in 2016 and 2017 in Shenyang and Kyoto at full heading for the EP and NEP genotypes with the AKI and LG5 backgrounds are shown in Fig.3.In Shenyang in 2016,the PH values for the EP and NEP genotypes with the AKI background were 85.6 and 103.1 cm,respectively;and for the LG5 background,the PH values for the EP and NEP genotypes were 95.0 and 119.9 cm,respectively. In Shenyang in 2017,the PH values for the EP and NEP genotypes with the AKI background were 87.7 and 103.6 cm,respectively;and for the LG5 background,the PH values were 97.3 and 115.0 cm for the EP and NEP genotypes,respectively. For Kyoto in 2016,the PH values for the EP and NEP genotypes with the AKI background were 81.2 and 103.0 cm,respectively;and the PH values for the LG5 background were 95.8 and 113.6 cm,respectively. In Kyoto in 2017,the PH values for the EP and NEP genotypes were 76.0 and 99.0 cm for the AKI background,respectively,and the values were 86.9 and 110.0 cm for the LG5 background,respectively.Therefore,the PH was lower for the EP genotype than for the NEP genotype for both backgrounds,and the lowest values were observed in AKI-EP across the four NILs.

Fig.3 Plant heights for the two genotypes with Akitakomachi (AKI) and Liaojing 5 (LG5) backgrounds grown at two locations in 2016 and 2017. EP,erect panicle;NEP,non-erect panicle. Bars mean SD (n=5).

3.3.Yield and yield components

The yield and yield components for the two genotypes with AKI backgrounds grown at the two locations over two years are shown in Table 2. The yields of the EP and NEP genotypes were 6.5 and 6.1 t ha-1in Shenyang in 2016,7.3 and 5.8 t ha-1in Kyoto in 2016,6.8 and 6.8 t ha-1in Shenyang in 2017,and 6.0 and 5.9 t ha-1in Kyoto in 2017,respectively. The mean yields were similar in 2016 and 2017. The mean yields in Shenyang and Kyoto were 6.5 and 6.3 t ha-1,respectively,with the yield being slightly higher (P＜0.05) in Shenyang than in Kyoto. The yield of AKI-EP was consistently higher than that of AKINEP in all four environments,with average values of 6.67 and 6.13 t ha-1,respectively.

The effect of the EP genotype on HI was significantly stronger than the NEP genotype (P＜0.001),whereas the effects of year and location were not significantly different.The EP genotype exhibited significant positive effects on PN (P＜0.001) and GNPM (P＜0.001) and significant negative effects on KGW (P＜0.001) and SSR (P＜0.001).However,the GNPP was significantly negative for AKIEP and positive for LG5-EP. The effects of the year on all yield aspects were significant for PN,GNPP,GNPM,KGW,and SSR,the values of which were lower in 2017 than in 2016,except for PN which was higher in 2017.The effects of location were significant for GNPP (P＜0.05)and KGW (P＜0.001),with higher GNPP values in Kyoto than in Shenyang,whereas KGW was higher in Shenyang than in Kyoto.

The yield and yield components for the two genotypes with the LG5 background grown in the two locations over two years are shown in Table 3. The yields for LG5-EP and -NEP were 6.77 and 6.78 t ha-1in Shenyang (2016),6.25 and 6.30 t ha-1in Kyoto (2016),7.04 and 6.96 t ha-1in Shenyang (2017),and 6.57 and 6.30 t ha-1in Kyoto (2017),respectively. The EP genotype had higher yields than the NEP genotype,but the difference was not significant. The mean yields for Shenyang and Kyoto were 6.89 and 6.35 t ha-1,respectively,which indicated higher yields in Shenyang than in Kyoto for plants with the LG5 background.

Table 2 Yield and yield components for the two genotypes with Akitakomachi (AKI) background grown at the two locations in 2016 and 20171)

The biomass of the EP genotype plants was significantly lower than the NEP genotype plants(P＜0.001),and HI was significantly higher for the EP genotype. The effect of location on biomass was significant,and biomass was higher in Shenyang than in Kyoto (P＜0.001). Regardless of year or location,PN,GNPP,and GNPM were significantly higher,whereas KGW and SSR were significantly lower for the EP genotype than for the NEP genotype.

In summary,the effect of the EP genotype was positive only for the AKI background. The effects of the EP genotype on PN and GNPM were relatively consistent and positive for both backgrounds,resulting in a positive effect on HI and a negative effect on KGW.

3.4.Effects of interaction between genotype and environment on yield and yield components

As shown in Table 2,the interaction between year and genotype (Y×G) in the AKI background significantly affected yield,biomass,KGW,and SSR (P＜0.01,P＜0.01,P＜0.01,andP＜0.001,respectively). In 2016,the effect of the EP genotype was stronger on yield but weaker on biomass. In 2017,the effect was weaker for both KGW and SSR. The interaction between location and genotype(L×G) significantly affected PN,GNPM,and SSR(P＜0.01,P＜0.01,andP＜0.001,respectively). For the EP genotype,the effect was stronger on PN and GNPM in Shenyang but weaker for SSR in Kyoto.

As shown in Table 3,the interaction between year and genotype (Y×G) for the LG5 background was only significant for KGW (P＜0.001). The effect of the EP genotype was weaker on KGW in 2017. The interaction between location and genotype (L×G) was significant for biomass,HI,PN,GNPP,and SSR (P＜0.01,P＜0.01,P＜0.001,P＜0.001,andP＜0.001,respectively). The effects of the EP genotype were stronger on HI and GNPP in Kyoto and on PN in Shenyang,with lower biomass and SSR in Kyoto and Shenyang,respectively.

3.5.Effectiveness of the EP genotype in relation to solar radiation and temperature

The interactions between genotype and environment exerted significant effects on yield (AKI background),biomass (both backgrounds),and HI (LG5 background).Matsushima’s V-shaped high-yielding theory states that the most influential period for yield determination is a 40-day interval from 15 days before heading to 25 days after heading (Matsushima 1969). Therefore,the meteorological data from this period were used to analyze the effectiveness of the EP genotype in relation to SR and TMEAN for 2016-2018 (Tables 4 and 5).

The effectiveness of the EP genotype with the AKIand LG5 backgrounds are shown in Tables 4 and 5,respectively. Effectiveness is represented by the ratio of the crop performance of the EP genotype relative to its NEP NILs. For the AKI background,the effectiveness of the EP genotype on yield and biomass was stronger in Kyoto (1.23 and 1.06,respectively) than in Shenyang(0.93 and 0.91,respectively). Similarly,for the LG5 background,the effectiveness of the EP genotype on yield and biomass was stronger in Kyoto (1.02 and 1.02,respectively) than in Shenyang (0.98 and 0.87,respectively). The effectiveness of the EP genotype on yield and biomass were stronger in 2016 and 2018 than in 2017.

Table 3 Yield and yield components for the two genotypes with Liaojing 5 (LG5) background grown at the two locations in 2016 and 20171)

Table 4 The effectiveness of the erect pancle (EP) genotype for yield and yield components,and their correlations with daily solar radiation and temperature for the two genotypes with the Akitakomachi background grown at the two locations in 2016,2017 and 20181)

Table 5 The effectiveness of the erect pancle (EP) genotype for yield and yield components and their correlations with daily solar radiation and temperature for the two genotypes with the Liaojing 5 background grown at the two locations in 2016,2017 and 20181)

With the AKI background,the variations in the effectiveness of the EP genotype on yield and biomass showed positive correlations with solar radiation and TMEAN for this period (Table 4;Fig.4) and was higher in Kyoto (18.0 MJ m-2d-1) than in Shenyang (16.8 MJ m-2d-1),and higher in 2016 (17.7 MJ m-2d-1) and 2018(20.9 MJ m-2d-1) than in 2017 (16.0 MJ m-2d-1). The effectiveness of the EP genotype on HI showed negative correlations with TMEAN and SR for the LG5 background,whereas they were positively correlated in the AKI background. The two backgrounds exhibited different responses to sink formation,such as PN. Effectiveness negatively correlated with TMEAN only for the LG5 background.

Fig.4 Effectiveness of the erect panicle (EP) genotype in relation to daily solar radiation in 2016 and 2017 at Shenyang,China and Kyoto,Japan and in 2018 at Kyoto with two backgrounds. AKI,Akitakomachi;LG5,Liaojing 5. HI,harvest index.

4.Discussion

In this study,two sets of NILs were investigated to determine their production capacity as a population and the effectiveness of the EP genotype was analyzed for different years and locations with two genetic backgrounds. The growth periods differed between genotypes,backgrounds,and locations. In Shenyang,the EP genotype for both backgrounds reached the heading stage earlier than the NEP genotype,whereas in Kyoto,the EP genotype with the AKI background reached the heading stage later than the NEP genotype. This difference in the effects of the EP genotype and phenology between the backgrounds may result in different effects of the interaction of the EP genotype and the environment between the two backgrounds on yield. A short growth period generally results in low biomass accumulation and poor development of yield organs,which offsets any positive effect of the EP genotype on yield (Takaiet al.2006). However,information on the possible effects of the EP genotype on the number of days until heading is unavailable,so further studies should be conducted to obtain these data.

The effects of the EP genotype on PN and GNPM were relatively consistent and positive for both backgrounds,resulting in positive effects on their respective HI values.Differences in HI between the EP and NEP genotypes were clearly evident: 0.56vs.0.50 for the AKI background and 0.52vs.0.49 with the LG5 background,respectively.These results,mostly for AKI-EP,were similar to those of 2018 (Appendix B). The effect of the EP genotype on biomass was negative for the LG5 background and neutral for the AKI background. This result is consistent with the findings of Tanget al.(2017),which showed that the biomass of the NEP type was significantly larger than that of the EP type. The improved HI in EP rice varieties has been emphasized through various comparisons(Hirookaet al.2018) and attributed to the effect of the EP genotype (Tanget al.2017). The findings of this study support and confirm those prior findings based on plant population observations of NILs. The results also indicated that the EP genotype had a positive effect on GNPM in many cases. The effect of the EP genotype was highly consistent with that of sink formation. In contrast,a negative effect of the EP genotype on the KGW led to a moderate effect of the EP genotype on yield. This finding might be a consequence of the lower or neutral biomass in the EP genotype compared to the NEP genotype. In addition,the EP genotype did not have a consistent positive effect on GNPP because of the different backgrounds,and the effect was evident only for LG5-EP. But the contribution of GNPP to grain yield was small for the EP genotype because the GNPP of LG5-EP was much larger in Kyoto and in 2016 only. These inconsistencies were due to the late and early heading of the EP genotype in 2016 at Kyoto and Shenyang,respectively,compared to the NILs of the NEP genotypes at both locations.

The average yields of the EP and NEP genotypes were 6.67 and 6.13 t ha-1,and 6.66 and 6.58 t ha-1for the AKI and LG5 backgrounds,respectively. The effect of the EP genotype on yield was positive only for the AKI background. These results might be attributed to the difference in PH between the backgrounds for the EP genotype. The AKI-EP PH was the shortest among the NILs,regardless of the location. However,the implications of PH for the yields of the EP genotypes need to be evaluated under different fertilizer regimes. The findings of this study support those of Makinoet al.(2021)and Idowuet al.(2022) because the yield of LG5-EP,which has longer flag leaf than AKI-EP,was no different from that of LG5-NEP,presumably because of its longer flag leaves. Makinoet al.(2021) showed that the short flag leaf of the EP genotype was a possible indicator of higher biomass accumulation after heading,compared to long flag leaves. Their findings agree with those reported by Idowuet al.(2022),who showed that AKI-EP had the shortest plant height due to its shortest flag leaf under different N fertilizer regimes.

However,the advantages of EP varieties depend on the ecological environment (Jinet al.2013). In this study,this dependence of the crop performance of NILs was confirmed. The daily solar radiation was higher in 2016 than in 2017,and higher in Kyoto than in Shenyang. As suggested by earlier studies,high solar radiation is advantageous for high rice yields (Penget al.2004). This conclusion is also consistent with the finding that prolonged insolation at the grain-filling stage is the main ecological factor contributing to high rice yield in northeastern China (Jinet al.2013). In this study,daily solar radiation was significantly correlated with the effectiveness of the EP genotype on yield and biomass with the AKI background (Fig.4-A). In Kyoto,the effectiveness of the EP genotype on yield was significantly higher in 2016 and 2018 than in 2017,and higher than in Shenyang. With the LG5 background,the effect of the EP genotype on yield exhibited no differences between the different years or locations,illustrating the lack of a significant correlation between solar radiation and the EP genotype. The effectiveness of the EP genotype on HI was either positive or negative with the AKI or LG5 backgrounds,respectively (Fig.4-C);high solar radiation negatively affected GNPP,and high solar radiation and temperature negatively affected GNPM and SSR with the LG5 background. The lack of consistency in the relationships between the effectiveness of the EP genotype on HI and the environment could have been due to the fact that the genotype on HI is only minimally influenced by the environment.

This study demonstrated positive correlations between TMEAN and effectiveness with yield and biomass because of the general association between high solar radiation and high day temperature. The maturity stage occurred later in Shenyang than in Kyoto;so consequently,TMEAN tended to be higher in Kyoto,and the environmental temperature was lower in Shenyang because of the differences in the climates of the two locations as well as the late harvest time in Shenyang. This difference also caused an apparent positive correlation between TMEAN and the effects of the EP genotype on yield and biomass.These findings are consistent with those of previous studies (Urairiet al.2016;Hirookaet al.2018). The high capacity for leaf photosynthesis may be a trait of the EP genotype owing to the higher radiation use efficiency(Idowuet al.2022). Therefore,the primary environmental factor that affects the effectiveness of the EP genotype is solar radiation.

5.Conclusion

In this study,the effect of the EP genotype on yield was found to be positive only with the AKI background because of the different phenological responses of the genotypes with the two different backgrounds. The effects of the EP genotype were relatively consistent and positive for both PN and GNPM,resulting in a positive effect on HI and a negative effect on KGW. The ratios of the performance of the EP genotype relative to the NEP genotype in terms of both yield and total biomass correlated positively with SR during the 40-day period around heading. These results suggested that the effect of the EP genotype is consistently positive for sink formation,but its effectiveness is conditional on source capacity,as it tends to be evident only in a high-radiation environment.

Acknowledgements

This study was supported by the Joint Funds of the National Natural Science Foundation of China (U1708231 and JSPS KAKENHI,26292013).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm